Analytics Date Grouping Data Discrepancies
Analytics Data Discrepancies
I just came across a new issue. I am running a report but I am getting different traffic for the days based on the date grouping I pull.
For Instance -
When I pull May 1st - May 31st here are the sessions for the first few days
5/1/2016 - 240
5/2/2016 - 326
5/3/2016 - 306
5/4/2016 - 297
5/5/2016 - 303
5/6/2016 - 332
5/7/2016 - 266
5/8/2016 - 229
5/9/2016 - 438
5/10/2016 - 395
When I pull May 1st - May 10th here are the sessions for the first few days
5/1/2016 - 239
5/2/2016 - 317
5/3/2016 - 304
5/4/2016 - 294
5/5/2016 - 298
5/6/2016 - 324
5/7/2016 - 257
5/8/2016 - 219
5/9/2016 - 320
5/10/2016 - 327
Most of the days aren't that far off, but as time goes the data becomes more varied (check May 9th and May 10th). As the month goes on the variation becomes greater.
We are pulling this data using an advanced segment that includes medium - organic and only new users. The segment is used the same across the board. And I get the same exact discrepancy every time I pull the data (it's repeatable).
I have checked sampling and changing the sampling rates doesn't affect it. I don't really know what else to check.
As I think through it, it doesn't seem like the segment should affect it because once the data is recorded it should just be there - and the fact that i'm getting the same differences every time I look at certain date groupings is also interesting.
Has anyone seen this before? Does anyone have any idea why this is the case?
Re: Analytics Date Grouping Data Discrepancies
"Do you have Filter Sessions or Filter Users selected in your Traffic source segment setting? It's easy to miss. This should be set to Filter Sessions so that you get a consistent result across date ranges.
Like +Tobias Kraeft said, Users segments are bound to the time range, and with a full-month time range you'd arrive at a different count than when you have only the first 10 days, since the Users could have had the "Organic" session at the end of the month, but they were a New User at the beginning of the month."